this meeting to enable at least the initial experiments to go 
forward. Dr. Anderson underlined that there were further reviews 
the experiments must undergo and that a delay in voting would in 
fact be amplified by causing further delays in other parts of the 
review process. 
Dr. Epstein said he was not sure whether an animal model was 
really required for the proposed experiment since it would not 
matter whether the results were positive or negative. However, 
he said he was disturbed by the fact that the RAC was being put 
into the position of passing on an experiment in which data had 
been deliberately withheld cn the basis of claiming that release 
of the data would jeopardize publication. He asked whether data 
presented today were on or off the record; and if they were off 
the record, how this affected their usage in the ultimate 
decision to approve or not approve the protocol. 
Dr. McGarrity noted that Mr. Lanman, the NIH Legal Advisor, was 
not present; but in Dr. McGarrity’ s opinion, the data were on the 
record insofar as the RAC was concerned. However, presentation 
of the data here did not constitute a "public disclosure." 
Dr. Anderson concurred and stated that all data presented at the 
meeting were on the record. 
Dr. McGarrity noted that references to the RAC as a "regulatory 
committee" were not totally accurate. In fact, by definition, 
the RAC is advisory to the Director of NIH but in reality could 
be construed to be a de facto regulatory body. 
Dr. Mclvor said he had three concerns: 
1. The ability of the PCR assay to detect virus 
sequences from the reinfused cells; 
2. The utilization of the S+/L- assay to detect 
replication competent virus; and 
3. The use of an animal model in which the N2 
vector could not infect the murine TIL. 
Dr. Mclvor said in the case of the two assays, he felt the 
Subcommittee should have a chance to evaluate the current status 
of these and to look at the data presented before making a 
recommendation. Further, on the animal model, he asked whether 
it were possible to extrapolate the findings in the murine model 
to the human because of the carefully controlled experimental 
approach to the model. 
Dr. Rosenberg explained there are four major differences in mouse 
and human TIL. All mouse TIL are CDS- [one class of lymphocytes 
defined by the presence of CDS cell surface markers], and CD4- 
Recombnant DNA Research Volume 13 
[24S] 
